This proposal describes iMStrong, a remote monitoring system that uses state-of-the-art technologies to identify the onset and duration of cocaine use in cocaine-dependent patients under treatment. iMStrong comprises three integrated components. The first are unobtrusive, wearable sensors that continuously record and wirelessly transmit physiologic measures (eg, increased electrodermal activity, skin temperature, and motion) of sympathetic nervous system (SNS) arousal. The second component is a smartphone that functions as a mobile computing device for data entry and storage. The third component is the iMStrong application (app) on the smartphone. The app will interpret the transmitted physiology data, looking for periods of intense SNS arousal that are characteristic of cocaine use. We will test iMStrong's ability to recognize cocaine use episodes over a 30 day period among patients (N=40) entering treatment for cocaine dependence. The iMStrong app will validate the identity of the user using commercial facial recognition protocols. The app will also identify periods of increased SNS arousal and request that participants annotate temporally dynamic social, behavioral, and environmental contexts surrounding these events. Three times a week we will obtain a quantitative urinalysis for cocaine. On a weekly basis, study staff will interview participants using the timeline follow back method to evaluate drug use episodes. After completion of the study we will evaluate user experience with iMStrong using summative interviews of participants. The specific aims are: 1) Technology deployment: To deploy and test a mobile technology for the remote identification of cocaine use episodes; and 2) Technology evaluation: To evaluate experience with the developed technology to understand the barriers and facilitators of technology utilization with a specific focus on detection of cocaine use as well as mechanisms of action that might underlie the quality of the patient-technology relationship. This study, which will obtain four different measures of cocaine use (real-time, continuous biophysical monitoring; annotations from mobile space; sequential urinalysis for cocaine metabolite; and self-report), is supported by preliminary data demonstrating our ability to detect cocaine use in natural environments. iMStrong will not only identify the onset and duration of episodes of cocaine use, it obtain formative information on temporal events surrounding relapse. iMStrong's collection of continuous data regarding the timing, location, contexts, and durations of cocaine use in individuals under treatment will advance drug abuse treatment research by providing a new capability to demonstrate efficacy of drug abuse interventions.